The De Havilland Comet Accidents
The de Havilland DH.106 Comet’s fall was literally marked by a fall from the sky-and only six months after it had entered service with BOAC in its original guise as the DH.106-1 on October 26, 1952.
Commanded by Captain R. E. H. Foote and carrying 35 passengers, the jetliner rotated at 112 knots from Rome’s Ciampino International Airport on its final, northbound leg to London, having originated in Johannesburg. Retracting its undercarriage and penetrating the night sky shortly before 19:00, it yawed to the left and began to stall. A corrective action, consisting of a climb angle decrease, failed to correct the anomaly. Unresponsive to the yoke movement, it slammed back on the ground, bouncing and ultimately remaining earthward, but with little distance in which to stop, it barreled over the runway’s end, plowing into a dirt mound and shedding its landing gear.
Sliding to a stop, it was able to permit all on board to escape without major injuries. Although the ruptured wings began to dislodge their fuel, the minor impact spared all of the sparks that could have ignited it.
The accident, the first since the type had entered service, was not ultimately attributed to design flaw or malfunction, because it was determined that the captain had tried to takeoff with an “excessive nose-up attitude,” as evidenced by the 650-yard scrape mark gouged in the runway created by the dragging tail.
While the aircraft itself was pronounced sound, the mishap only six months after it entered service did little to quell the opinions of those who believed that pure-jet technology had been prematurely introduced. Nevertheless, could it have been a forewarning of events to come-that confirmed these beliefs? Four months later, on March 3, 1953, the question would once again be posed.
Piloted by Captain Charles Pentland, a Comet 1A operated by Canadian Pacific Air Lines named “Empress of Hawaii,” was undertaking its delivery flight from London to Sydney with intermediate stops in Karachi and Singapore. Ten de Havilland and Canadian Pacific representatives were on board. The water methanol-injected version was slated to inaugurate the carrier’s Sydney-Honolulu service once it had reached its final destination.
Like the BOAC Comet 1 in Rome, the aircraft attempted an acceleration run in Karachi, but it produced so little lift, that it could not achieve airborne status. Lacking sufficient distance in which to stop, it, too, passed over the runway’s end, plowing through a fence and rupturing its full fuel tanks, which ignited into sky-extending flames. Although it ultimately came to rest 80 yards from the concrete, a second explosion claimed the lives of all eleven on board.
The two strikingly similar accidents were carefully reviewed. It seemed unlikely that both pilots had made the same error, leaving investigators to reconsider an aircraft design flaw.
It was subsequently discovered that, because of the aircraft’s symmetrical wing, that it had to rotate at a very small angle in order for it to generate proper lift and anything above it ironically retarded it-or reversed the very purpose for which it was designed. Integral to this phenomenon was the fact that the engine intakes, dependent upon this angle, required a precise angle-of-attack so that the accepted air could pass through the turbine sections and produce their optimum thrust. Any angle equal to or greater than nine degrees produced a stall and any equal to or greater than 11.5 degrees caused the tail to scrape on the ground. These conditions, along with the fact that no cockpit instrumentation existed to indicate this excessive rotation and that both the BOAC and Canadian Pacific incidents occurred at night without outside horizon references, were ultimately determined as the true causes of the mishaps.
Remedially, the aircraft’s wing leading edge was redesigned, incorporating a downward curve and the upper portions of the engine intake ducts were retrofitted with similar curved contours, both ensuring safer rotations at a wider range of angles. Since the type’s lift potential increased, so, too, did its payload-in this case, by 2,000 pounds.
Air France was the first carrier to operate the type in modified form.
Whether this could be considered a “design flaw” or a temperamental characteristic of the aircraft was debatable, but it did alert pilots to the fact that pure-jet aircraft had to be precisely flown “by the book.” Speed had to be respected, since it was achieved with compromises, and failures could result in life-losing payments-which is what occurred on the one-year anniversary of the type’s service inauguration.
Piloted by Captain Maurice William Haddon, who had logged 587 hours in type, First Officer Robert Strange, Flight Engineer Albert Gilmore, and Radio Officer Alfred Wood, the BOAC Comet 1 operating between London and Singapore that day and registered G-ALYV had been warned of severe thunderstorms in the Calcutta area, particularly at 16:30, the scheduled departure time for its leg to Delhi. Now in the midst of its monsoon season, India, not surprisingly, was regularly racked with driving rain, pounding winds, and electrical lightning and thunderstorm activity.
Touching down in Calcutta’s Dum Dum Airport at 15:10, the jetliner dislodged its passengers and was refueled. While the next leg was predicted to entail clear skies, local thunderstorm activity continued to rage. Nevertheless, the captain elected to proceed with the flight.
Executing its acceleration roll on Runway 19-Left, the Comet divorced itself from Indian soil at 16:19 with 37 passengers, initiating a right bank and trimming itself for a 1,000-fpm climb. Its assigned altitude was 32,000 feet.
Passing through a thunderstorm cell and hammered by 50-knot winds, the aircraft was tossed about like a toy plane, unable to transmit to Delhi, with which it had established contract. It was 16:36.
Attempting to gain control, Captain Haddon disengaged the autopilot and reduced power to the recommended turbulence penetration speed, but it was futile: gripped by the clutches of a downdraft, it was forced into a nose-down attitude. Losing altitude, but gaining airspeed, its profile was counteracted with elevator deflections, but these proved of little use. Creating beyond-design tolerance, their force only succeeded in snapping off the tail, and the downward pressure on the wings equally caused them to dislodge themselves from their mountings.
One sliced through the rudder, the only remaining axis control surface, like a knife and the other gouged through the fuselage like a saw. The mangled, ruptured fuel tanks spilled conflagration-creating flames, which clung to whatever of the original structure still existed, and the jetliner lit a path through the thunderstorm cell as it meteored to the ground and dug such a deep ravine upon impact that local residents initially thought that an earthquake had erupted from within the earth. The fuel brimming tanks lit such a fire that even the monsoon’s deluge failed to douse it. All 43 on board, needless to say, perished.
Ironically, no one was aware of the events. Calcutta, having had difficulty contacting the aircraft because of the static, assumed that it had spoken with Delhi instead, and it had, in fact, attempted to do so. An hour after the crash, it was Delhi that reached out to Calcutta to determine the aircraft’s’ location. It was only then that they realized that neither had been in contract with it.
The wreckage, subsequently discovered 22 miles northwest of the airport, indicated that it had sustained structural failure as a result of the intense thunderstorm activity.
The probably cause was determined as “structural failure of the airframe during flight through a thunder squall. In the opinion of the Court, the structural failure was due to overstressing which resulted from either severe gusts encountered in the thunder squall or overcontrolling or loss of control by the pilot when flying through the thunderstorm.”
But, again, no design flaw was suspected or uncovered.
Nevertheless, the latest accident, which took Comet fatalities to 54 in only a year of operation, once again cast doubts about the reliability and safety of pure-jet travel, and skeptics maintained that its speed had eclipsed technology. Regardless of their accuracy, their convictions were again proven at the beginning of the new year-or on January 10, 1954.
Operating between Singapore and London on that day, a BOAC Comet 1 registered G-ALYP and under the command of Captain Alan Gibson, made its scheduled refueling stop in Rome, where a British European Airways (BEA) pilot joined the 28 passengers already on board. Routinely rotating at 10:31 and disengaging itself from Italian soil at a 105-knot speed, it trimmed itself for a 19-minute ascent that took it through and over majestic cloud tops to the morning blue of 26,000 feet.
A shallow climb over the coast of Italy would take it another 10,000 feet to its 36,000-foot assigned altitude and its flight plan over France to complete the final leg of its multi-sector trip to London. But it would never arrive there.
A bomb-like explosion ripped the cabin apart, rupturing a wall, which caused the roof to collapse and be ejected from its mountings. The wings, reduced to three sections, snapped off like twigs. Almost vertically diving, the remainder of the fuselage plunged into the sea off the island of Elba with such ferocity that the nose and tail sections, unable to withstand the g-forces, were torn off the structure before the impact and fell into the water at considerable distances from the rest of the wreckage.
If there had been any survivors, perhaps some could have revealed the mystery that caused the explosion and that would ultimately surface on its own. Reduced to just bits and pieces, the aircraft, which had launched the jet age when it operated the inaugural flight in 1952, would have to be reassembled before any probable cause could be identified. Ironically, it would now be one of the two that would terminate it.
The latest-and far deadlier-crash, needless to say, cast doubts about the Comet’s design integrity, its premature entry into the jet age, and the reasons behind the other three. Perhaps the aircraft was not as strong or as safe as previously believed, it could only be wondered. Perhaps it regularly plied the skies with some secrecy that the accidents were just beginning to reveal-if it could only be determined what it was. Even BOAC’s confidence, which had proudly placed the first pure-jet flagship into service and which therefore almost became symbolic of it, began to evaporate.
Thirty-seven hours after the latest mishap, at a time when all of the Comets in its fleet were on the ground somewhere in the world, it issued the following statement.
“As a measure of prudence, the normal Comet passenger services are being temporarily suspended to enable minute and unhurried technical examination of every aircraft in the Comet fleet to be carried out at London airport.”
Four DH.106-1s, currently away from their home base, were ferried back without passengers at low cruise altitudes.
What was happening? The pure-jet airliners, once considered trail blazers and innovative leaders that would prove to the world that their superior technology, speed, and comfort would rapidly outdate all piston and turboprop transports, had become deathtraps. Perhaps de Havilland’s vision had been too far-reaching.
Several theories as to why the aircraft had crashed surfaced. One focused on fatigue, which had weakened the fuselage structure and caused the pressurized cabin to explode. But this was quickly discounted because of the extensive precertification tests the aircraft had been subjected to and its water tank immersion would certainly have revealed such a flaw. Another centered around sabotage-that is, that a bomb had been clandestinely planted somewhere on the airplane to tarnish the reputation and hinder de Havilland’s lead over US manufacturers.
Boeing, which was about to fly the 367-80, which itself would serve as the prototype of its first 707 jetliner, was, to the contrary, eager to learn about the Comet’s deficiencies, if any existed, so that it could incorporate its lessons in its own design.
Although the commercial aviation industry was competitive, it was also collaborative when it came to issues of safety and sought to remove all barriers to shared information concerning this aspect.
What became particularly mystifying and frustrating was the fact that the Comet’s flight test program had been more extensive that it needed to have been and no flaw or deficiency ever surfaced during it. Of course, no program, regardless of its extensiveness, could ever truly simulate actual airline service conditions.
Because de Havilland was adamant about not permitting the Comet to re-enter service if it carried a design flaw with it, it initiated the most extensive and expansive wreckage recovery, investigation, and reconstruction process the world had ever known. The Mediterranean Sea was dragged and dredged, during which every minuscule piece of the aircraft that could be found was collected. But it was not until a month and two days after the accident, or February 12, that the first piece was found seven miles off of Cape Calamita.
Because no significant number of pieces could be located, however, the effort, despite its merit, yielded no clues. Nevertheless, careful scrutiny of the seven other existing Comets resulted in the incorporation of 50 modifications which it was believed would have compensated for any design weakness.
Re-pronounced airworthy, the aircraft was cleared for service re-entry with BOAC, which occurred on May 23, 1954. The carrier had intermittently lost 50,000 pounds Sterling per day since it had been grounded. That it experienced full load factors with that very first, multi-sector flight from London to Johannesburg, indicated that public confidence had not been lost.
While de Havilland had not found the definitive cause of the Elba accident, it believed that the extensive modifications it had since made produced a more refined aircraft. What it did not know is that it let it re-enter the sky with a design flaw that would shortly re-reveal itself.
Waiting, like a ticking time bomb, to do so, the Comet fleet once again straddled the globe, carrying the passengers who would become victims to it. And it did not take long to do so. In fact, the interval was all of two weeks.
Piloted by Captain Willem Mostert and First Officer Barent Grove and served by Steward Jacobus Kok and Stewardess Pamela Reitz, a BOAC Comet 1 leased to South African Airways and registered G-ALYY, carried 14 passengers on its scheduled flight between London and Johannesburg on April 8, 1954. Landing in Rome-Ciampino the previous evening with a malfunctioning fuel gauge, it was delayed while a new one was flown in from London, not permitting its own departure until 18:32.
Subsequently climbing through 28,000 feet 26 minutes later, now bound for Cairo, it made its routine radio call.
“I am bound for Cairo, where my estimated time of arrival is 21;20,” reported Mostert. “I am ascending to 35,500 feet.”
The transmission was his last. As had occurred with aircraft G-ALYP, an explosion reduced the Comet to shreds, which were catapulted earthward, leaving them to submerge and scatter below the Bay of Naples off of Stromboli Island in the Tyrrhenian Sea. It was not until a BEA AS-57 Ambassador, overflying the area that what remained of the aircraft was spotted.
Not wasting another minute, the Minister of Transport and Civil Aviation rescinded the aircraft’s airworthiness certificate and grounded all the remaining Comets nine hours after the latest accident, permanently sealing its fate. The DH.106-1 would never again carry a single fare-paying passenger. Promise had been pummeled. Initial victory had been turned into ultimate defeat. And, in the next few days, the sound of slower-flying piston airliners once again filled the sky.
The jet age had been glorious, but brief. Had it come too fast? Had it been premature? Had de Havilland’s design engineers been too optimistic? Would it ever return? The answers lay in the widely scattered wreckage of the two aircraft off the coast of Italy that had claimed 110 lives. Now nothing would be spared to ascertain the answer-to discover the mystery that had been revealed as clues, once investigators could piece them together.
Because of the Comet’s advanced technology and the lack of commercial jet engine experience, there was little upon which to base the investigation, compounding the already difficult to impossible task of dredging the bottom of the ocean for evidence that took form as wreckage. It was sometimes wondered if the aircraft had entered a flight realm that even engineers could not fully understand or appreciate. After all, its speed and altitude realms were foreign to piston airliner operation.
Commonalities between the last two accidents were identified. In this case, both had just departed from Rome and both had been subjected to inflight explosions of an unknown origin.
Professor Antonio Fornari, Director of the Institute of Forensic Medicine of the University of Pisa, examined some of the victims of the BOAC crash and concluded that all had died “by violent movement and explosive decompression,” a fact concluded by the profundity of skull fractures and blown lungs sustained before death had occurred. The cause, it could only be concluded, was explosive decompression.
Although precertification tests indicated that the airframe was rated for at least a 10,000-hour failsafe structural life, the method of testing this would be redeployed and, in the process, challenge the validity of the explosive decompression theory. A 112-foot-long, 20-foot-wide, 16-foot-deep steel tank, along with its own reservoir, constructed to simulate inflight pressure, housed the fuselage of aircraft G-ALYU, whose wing openings were secured with pneumatic tubes. Water, pumped into both the tank and the fuselage, created a 8.25 pound per square inch pressure, and the wind and turbulence routinely experienced in the air were simulated by deflecting the protruding airfoils with hydraulic pistons. A three-hour flight cycle was recreated during a five-minute interval.
Wreckage from both accidents was, in the meantime, retrieved from the sea. Numerous sections, pieces, and parts belonging to aircraft G-ALYP were identified, tagged, and transported to London, where they were attached to a 90.1-foot-long, ten-foot-wide wooden frame at the Royal Aircraft Establishment in Farnborough. Seventy-five percent of the Comet’s airframe was ultimately recovered.
The water tank submerged aircraft had accumulated 9,000 simulated hours in the air by this time-and finally provided a clue to the mystery that had brought down two aircraft. The skin at the corner of a window had failed and ruptured, propagating to the fuselage and producing an eight-by three-foot hole through which an explosion, similar to that of both the BOAC and SAA aircraft, would have escaped if it had not been for the outside, equalizing water. Metal fatigue in both the simulated and actual Comets had clearly been the culprit and cause behind the explosive decompressions.
Based upon these events, it was concluded that the aircraft’s construction, originally believed to have been sufficient based upon earlier piston airliner design, was inadequate for the forces of pressurization that were double those exerted on its propeller predecessors at altitudes that were some 15,000 feet higher, rendering its 8,000 to 10,000-hour stress life calculation inaccurate.
Additional evidence to support this conclusion arose from a portion of G-ALYP’s roof, which was located off of the island of Elba by trawlers on August 12 and contained a similar sky window where the initial structural failure had occurred. When the weakness it created had spread to a wider area, it had succumbed to the cabin pressurization’s enormous forces and gave way, exploding.
Paint scratches covering the outside fuselage were detected on the left wing, indicating that the side had blown out toward it. The fact that consistent pressurization cycles, necessitated by the aircraft’s multiple daily takeoffs and landings, had only served to further weaken the structure because of the expanding (pressurization) and contracting (depressurization) forces exerted against it. The phenomenon was known as “hoop stress.”
De Havilland designers were ignorant of the stresses exerted upon an aircraft flying at 40,000-foot altitudes, because piston types regularly occupied flight levels that were half of them and no other, comparable jetliner existed with which to have compared data and experience. The Comet was, in essence, the aircraft which supplied this initial data-in this case, after fatal experience. If it had equaled earlier, propeller airliner operational realms, its service life would have been double what was calculated, but its speed, range, and economy mandated something far more radical
The DH.106’s design flaw was, in the end, not a result of whether the aircraft could withstand the tremendous pressurization forces exerted against it, but how long it could withstand them as a result of its vastly different operational parameters. Because it was subjected to some 16 pressurization cycles during a single London-Singapore flight, its stress life was quickly exhausted.
This costly and fatal experience, which never surfaced during precertification tests, was ultimately extracted from the wreckage of the two downed airliners and incorporated in every subsequent aircraft-Comet or otherwise.
As a result of this experience, it could no longer be debated whether the jet age had arrived too soon, since the lack of experience could well have resulted in the same design flaw where and whenever the first jetliner had been produced, even by other manufacturers. But, that the de Havilland Comet had been first, had represented a technological leap, and had made a plunge into the unknown, could not be disputed, despite the results.
The accident inquiry, taking placing in the Assembly Hall of London’s Church House and requiring 22 days to complete, resulted in a 48-page report published on February 12, 1955. It did not assign blame to any single person or group nor did it highlight the design’s failure and flaw. Instead, it said, in part, “There were men with foresight and vision enough to recognize that it was not the bold concept of the Comet which had failed, but metallurgy, the knowledge of which had been outpaced by the technological advances of aeronautical design.”
The DH.106 Comet’s unprecedented accidents and their subsequent investigations ultimately revealed the unknown, unexperienced mysteries of high-speed and -altitude pure-jet flight, and their knowledge was unprecedently distributed to the world’s aircraft manufacturers, enabling them to incorporate its lessons in every future airliner.
Although no doubt as to the definitive cause of the early Comet 1 accidents remained, future incidents were still carefully scrutinized for potential explosive decompression reasons. Seven fatal ones occurred during the ten-year period between 1958, when the Comet 4 re-entered service, and 1968.
The first of these took place on August 27, 1959 when an Aerolineas Argentinas Comet 4, registered LV-AHP, crashed near Asuncion, Paraguay, killing two, but its probable was its failure to maintain its minimum descent altitude (MDA) during an instrument, nondirectional beacon (NDB) approach.
The event was followed by mishaps to Comets operated by BEA; EgyptAir, which had originally been known as United Arab Airlines; and King ibn Saud’s private aircraft. Five years-namely, 1960, 1964, 1965, 1966, and 1968-were accident-free. The last crash in this period, which claimed the greatest number of lives, occurred on October 12, 1967 when a BEA Comet 4B, registered G-ARCO and operating the London-Athens-Nicosia-Cairo route, exploded over the Mediterranean Sea near Rhodes, Greece, killing all 66 on board.
Reminiscent of the South African Airways Comet 1 explosive decompression on April 8, 1954, it served to rekindle doubts, since its wreckage was widely scattered on the ocean floor, awaiting retrieval. But a passenger seat and other debris ultimately pointed to the internal explosion of a high detonation device and subsequent autopsies confirmed this suspicion. The motivation, it was conjectured, was the attempt to kill General Grivas, who was on board and was once the leader of the Ethniki Organosis Kyprion Agoniston (EOKA), a Greek Cypriot nationalist paramilitary organization that fought a campaign for the end of British rule in Cyprus and for its eventual union with Greece.
None of these accidents was attributed to explosive decompression resulting from insufficient fuselage skin gauges.